Economy

Highlights Fed: The Fed is preparing markets for a taper announcement in Q4 of this year. But we don’t see asset purchase tapering as a catalyst for higher bond yields. Rather, bond yields will move higher as the employment data continue to come in hot. Job growth will be strong enough to reach the Fed’s definition of maximum employment by the end of 2022, and the fed funds rate will rise more quickly than is implied by current market expectations. Duration: The 10-year Treasury yield will reach a range of 2% to 2.25% by the time the Fed is ready to lift rates, near the end of 2022. Strong employment data will catalyze the next significant jump in bond yields, but this may not happen until Q4 of this year. The spread of the delta COVID variant could limit the pace of hiring during the next month or two, and bond market positioning may need to turn more bullish before yields can rise. Labor Market: After July’s strong employment report, we calculate that average monthly nonfarm payroll growth of 431k is required to reach the Fed’s “maximum employment” liftoff criteria by the end of 2022. Feature Chart 1A Tapering Announcement Is Coming It’s finally time to talk about tapering. Several Fed governors and regional presidents made media appearances last week, each one presenting a timeline that sets up a tapering announcement before the end of this year. Federal Reserve Governor Christopher Waller: I think you could be ready to do an announcement by September. That depends on what the next two jobs reports do. If they come in as strong as the last one, then I think you have made the progress you need. If they don’t, then I think you are probably going to have to push things back a couple of months.1 St Louis Fed President James Bullard: I don’t think that we need to continue with these purchases now that we’ve got new risks on the horizon and possibly inflation risks on the horizon. […] What I think we should do here is start sooner and go faster and get finished by the end of the first quarter of next year. We don’t really need the purchases anymore.2 Dallas Fed President Robert Kaplan: As long as we continue to make progress in July (jobs) numbers and in August jobs numbers, I think we’d be better off to start adjusting these purchases soon. Doing so gradually, over a time frame of plus or minus about eight months, will help give ourselves as much flexibility as possible to be patient and be flexible on the fed funds rate.3 Fed Governor Lael Brainard presented the most detailed description of what it will take for the Fed to start paring its asset purchases.4 Since December, the Fed’s criteria for tapering has been “substantial further progress” toward its employment and price stability goals. In December, nonfarm payrolls were about 10 million below pre-pandemic levels (Chart 2A). In her speech, which was given prior to the release of July’s jobs report, Brainard noted that if employment grows at the same rate in Q3 as it did in Q2, then “about two-thirds of the outstanding job losses as of December 2020” would be made up by the end of 2021. That figure rose to 71% after July’s strong jobs number (Chart 2B). Chart 2AConditions For Tapering Chart 2BDefining "Substantial Further Progress" In other words, as long as employment growth stays solid – in the 500k/month range – then the Fed will be well over 50% of the way toward its maximum employment goal by the end of this year. This would certainly count as “substantial further progress”. Our expectation is that Q3 jobs growth will be strong enough for the Fed to make an official taper announcement in Q4, with the actual tapering starting in January 2022.5 There is an outside chance that the Fed will rush to start tapering earlier, but only if long-dated inflation expectations rise to well above the Fed’s target range (Chart 2A, bottom panel). As for market impact, we don’t expect the tapering announcement to move markets all that much. First, we mainly care about asset purchase tapering because it could signal that the Fed intends to move more quickly toward rate hikes (Chart 1). This is the concern that prompted the 2013 taper tantrum. This time around, however, the Fed has tied liftoff to explicit employment and inflation criteria. This forward guidance significantly weakens the signaling power of any tapering announcement. Second, surveys indicate that market participants already anticipate that tapering will start in early-2022 (Tables 1A & 1B). In other words, a Q4 taper announcement shouldn’t be that much of a shock to expectations. Table 1ASurvey Of Market Participants Expected Fed Timeline Table 1BSurvey Of Primary Dealers Expected Fed Timeline Interestingly, Fed Vice-Chair Richard Clarida did manage to shock markets with his speech last week, but only because he went further than just a discussion of tapering. Specifically, Clarida articulated his expected timeline for lifting interest rates: Chart 3Median FOMC Forecasts While, as Chair Powell indicated last week, we are clearly a ways away from considering raising interest rates and this is certainly not something on the radar screen right now, if the outlook for inflation and outlook for unemployment I summarized earlier turn out to be the actual outcomes for inflation and unemployment realized over the forecast horizon, then I believe that these three necessary conditions for raising the target range for the federal funds rate will have been met by year-end 2022.6 What are the economic forecasts that Clarida says would meet the conditions for liftoff by the end of 2022? It turns out that they are very close to the FOMC’s median projections (Chart 3). The Fed’s forecast calls for 3% core PCE inflation in 2021, falling to 2.1% in 2022 and 2023. The Fed also sees the unemployment rate falling to 4.5% by the end of this year, 3.8% by the end of 2022 and 3.5% by the end of 2023. Clarida said that he views this forecast as consistent with overall employment returning to its pre-pandemic levels by the end of 2022. We think Clarida’s expected timeline is reasonable. The Appendix at the end of this report presents different scenarios for when the Fed’s “maximum employment” liftoff condition might be met. We estimate that average monthly nonfarm payroll growth of 431k will get us to maximum employment by the end of 2022, in time for early-2023 liftoff. At least so far, monthly nonfarm payroll growth is tracking well above the 431k threshold. If we compare our (and Clarida’s) forecast to market prices, we conclude that market rate expectations are too low. The overnight index swap curve is priced for Fed liftoff in January 2023 but for not even three 25 basis point rate hikes in total by the end of 2023 (Chart 4). This seems too low if the Fed’s liftoff criteria are in fact met by the end of 2022, as is our expectation. Chart 4Rate Expectations Bottom Line: The Fed is preparing markets for a taper announcement in Q4 of this year. But we don’t see asset purchase tapering as a catalyst for higher bond yields. Rather, bond yields will move higher as the employment data continue to come in hot. Job growth will be strong enough to reach the Fed’s definition of maximum employment by the end of 2022, and the fed funds rate will rise more quickly than is implied by current market expectations. Timing The Move Higher In Yields Our expectation for a return to maximum employment by the end of 2022 implies that bond yields will be significantly higher by then. Specifically, we expect that both the 5-year/5-year forward Treasury yield and the 10-year Treasury yield will be in a range between 2% and 2.25% by the time of the first rate hike (Chart 5). The 2% to 2.25% range is consistent with survey estimates of the long-run neutral fed funds rate. But a big question remains over the timing of the next move higher in yields. Are bond yields poised to jump higher immediately? Or will they remain low for the next few months and move up only in 2022? Our sense is that the catalyst for the next significant jump in bond yields will be surprisingly strong employment data. There is widespread consensus that inflation will be close to the Fed’s target (if not higher) by the end of 2022, but recent concerns about labor supply have increased the uncertainty around employment projections. Ultimately, we think that labor supply constraints will ease and that the unemployment rate will catch up to levels implied by different labor demand indicators (Chart 6). However, this may not happen during the next month or two. Chart 5A Target For Long-Dated Yields Chart 6Labor Demand Is Strong The spread of the Delta coronavirus variant has just started to ramp up in the United States (Chart 7). The UK’s experience with the variant shows that vaccination significantly limits the number of hospitalizations and suggests that economic lockdowns can be avoided. However, it took about one month for the UK’s new case count to peak once the variant started spreading. A similar roadmap could lead to hiring delays in the US during the next month or two, at least until the new case count starts to fall and concerns abate. From a market technical perspective, we also note that bond market positioning remains significantly net short and that bond market sentiment is less bullish than is often the case at major inflection points (Chart 8). This is not the ideal technical set-up for a large immediate jump in bond yields. Chart 7Delta Is A Near-Term Risk To Hiring Chart 8Positioning & Sentiment Bottom Line: The 10-year Treasury yield will reach a range of 2% to 2.25% by the time the Fed is ready to lift rates, near the end of 2022. Strong employment data will catalyze the next significant jump in bond yields, but this may not happen until Q4 of this year. The spread of the delta COVID variant could limit the pace of hiring during the next month or two, and bond market positioning may need to turn more bullish before yields can rise. Appendix: How Far From “Maximum Employment” And Fed Liftoff? Chart A1Defining “Maximum Employment” The Federal Reserve has promised that the funds rate will stay pinned at zero until the labor market returns to “maximum employment”. The Fed has not provided explicit guidance on the definition of “maximum employment”, but we deduce that “maximum employment” means that the Fed wants to see the U3 unemployment rate within a range consistent with its estimates of the natural rate of unemployment, currently 3.5% to 4.5%, and that it wants to see a more or less complete recovery of the labor force participation rate back to February 2020 levels (Chart A1). Alternatively, we can infer definitions of “maximum employment” from the New York Fed’s Surveys of Primary Dealers and Market Participants. These surveys ask respondents what they think the unemployment and labor force participation rates will be at the time of Fed liftoff. Currently, the median respondent from the Survey of Market Participants expects an unemployment rate of 3.5% and a participation rate of 63%. The median respondent from the Survey of Primary Dealers expects an unemployment rate of 3.7% and a participation rate of 63%. Tables A1-A4 present the average monthly nonfarm payroll growth required to reach different combinations of unemployment rate and participation rate by specific future dates. For example, if we use the definition of “maximum employment” from the Survey of Market Participants, then we need to see average monthly nonfarm payroll growth of +431k in order to hit “maximum employment” by the end of 2022. Table A1Average Monthly Nonfarm Payroll Growth Required For The Unemployment To Reach 4.5% By The Given Date Table A2Average Monthly Nonfarm Payroll Growth Required For The Unemployment To Reach 4% By The Given Date Table A3Average Monthly Nonfarm Payroll Growth Required For The Unemployment To Reach 3.5% By The Given Date Table A4Average Monthly Nonfarm Payroll Growth Required To Reach “Maximum Employment” As Defined By Survey Respondents Chart A2 presents recent monthly nonfarm payroll growth along with target levels based on the Survey of Market Participants’ definition of “maximum employment”. This chart is to help us track progress toward specific liftoff dates. For example, if monthly nonfarm payroll growth continues to print at the same level as last month, then we could anticipate a Fed rate hike by June 2022. Table A2Tracking Toward Fed Liftoff We will continue to track these charts and tables in the coming months, and will publish updates after the release of each monthly employment report. Ryan Swift US Bond Strategist rswift@bcaresearch.com Footnotes 1 https://www.bloomberg.com/news/articles/2021-08-02/waller-says-strong-job-reports-may-warrant-september-taper-call?sref=Ij5V3tFi 2 https://www.stlouisfed.org/from-the-president/video-appearances/2021/bullard-washington-post-inflation-tapering 3 https://www.reuters.com/business/finance/exclusive-feds-kaplan-wants-bond-buying-taper-start-soon-be-gradual-2021-08-04/ 4 https://www.federalreserve.gov/newsevents/speech/brainard20210730a.htm 5 Please see US Bond Strategy / Global Fixed Income Strategy Special Report, “A Central Bank Timeline For The Next Two Years”, dated June 1, 2021. 6 https://www.federalreserve.gov/newsevents/speech/clarida20210804a.htm Recommended Portfolio Specification Other Recommendations Treasury Index Returns Spread Product Returns

Highlights Economy – A range of economic and fundamental indicators are at such high levels that deceleration is inevitable: US growth will peak any day if it hasn’t done so already. Markets – Financial markets typically pay closer heed to direction than level: All else equal, we prefer direction to level as well, but levels are likely to remain elevated for a while even as deceleration takes hold, and investors should take that into account when assessing the outlook. Strategy – Remain overweight equities and credit in multi-asset portfolios: Risk assets are likely to continue to generate positive excess returns over Treasuries and cash despite moderating growth. Feature COVID-19’s arrival ushered in a wave of extremes in monetary and fiscal policy measures, economic data and financial markets. Everywhere investors look, data series are at unusually outlying levels. Inflation pressures are more intense than they have been in decades, as measured by consumer price indexes and a range of business surveys. Household net worth has advanced at its fastest-ever five-quarter pace despite the record setback that began the pandemic, S&P 500 earnings growth has demolished analyst expectations over the last five quarters, the federal government has injected a head-spinning amount of fiscal stimulus into the economy and the Fed has done all it seemingly could to cushion the pandemic’s economic blow. Much of the growth has resulted from Herculean stimulus measures that cannot be maintained on a rate-of-change basis. The slowdown in fiscal and monetary thrust implies that economic growth, along with several other series that are viewed as significant financial market drivers, will soon peak if they haven’t already. The looming deceleration has kindled a recurring debate among BCA researchers: What matters most for financial markets, level or direction? The answer to the most challenging questions in markets and economics is often “it depends,” and that’s the way we view the level-versus-direction debate. We’d position a portfolio based on direction if key series were just breaking above or below trend levels with robust momentum, but it’s a more nuanced decision when they are slowing from exceedingly high levels and a modestly decelerating pace should have them still sitting well above trend this time next year. Our view, then, is that the interaction between level and direction will drive markets going forward. Given that we have cited a range of levels in support of our bullish stance, however, it is prudent to ask how good might be too good for reliably mean-reverting series. We therefore examine the empirical record of how S&P 500 returns have interacted with the level and direction of the unemployment rate, earnings-per-share growth, and interest rates. We conclude that the humble level matters as well as the more celebrated rate of change and that deceleration will not spell the end of the equity bull market. The Unemployment Rate We used the unemployment rate as a proxy for the impact of macroeconomic changes on S&P 500 returns. While the unemployment rate is quite variable from month to month, it tends to follow a clear pattern over longer periods of time, rising very rapidly to cyclical peaks before meandering its way to cyclical troughs. Over the series’ 73-year history, there have been eleven complete rising phases and it is currently in its eleventh declining phase (Chart 1). Owing to unemployment’s established pattern – it takes the elevator up and the stairs down, flipping equity indexes’ pattern on its head – the eleven rising phases have spanned 30% of the nearly 900 months while the falling phases currently total 70% of them. Chart 1Unemployment Takes The Elevator Up And The Stairs Down A simple compilation of one-month forward S&P 500 returns based on the level of the unemployment rate has a clear theme – stocks do well when the rate is at least one standard deviation above the mean (about 7.4% or higher) and poorly when it is one or more standard deviations below it (about 4.2% or lower) (Chart 2, left side). A compilation based on the month-to-month direction of the unemployment rate – up, down or unchanged – also favors rising unemployment, though it is unclear what investors should conclude from the fact that rising and falling both outperform unchanged (Chart 2, right side). Chart 2The S&P 500 Is Sensitive To Anticipated Turns In Unemployment We think the analysis is much improved if the unemployment rate is combined with its direction as indicated by the cycle phase. The interaction of level and phase provides more information than the simple message that high unemployment is good for equities and low unemployment is bad. Applying the rising or falling unemployment rate phase to the ranges shown in Chart 2, we find that direction matters quite a lot within four of the five ranges, where the annualized return differs by thirteen to sixteen percentage points based on the underlying trend (Table 1). Table 1Level And Direction Tell The Most Compelling Story Equities are just coming off their bottom, on balance, when the unemployment rate exceeds a standard deviation above its mean and is still rising. Direction is everything when the rate is below its mean (5.8%). When it’s falling, there’s plenty of money to be made in an expanding economy before the Fed has designs on removing the punch bowl, though once the rate is a standard deviation below the mean (4.2% or lower), the equity top is near. Once the unemployment rate rises off the bottom, even though it’s still at an unusually low level, the equity tide has already begun to go out. Losses are in store until the rate gets back above the mean, signaling future improvement. Chart 3Up, Up And Away It is important to recognize that we can only demarcate the unemployment rate’s phases in retrospect. There is no telling with certainty in real time how far a nascent trend will go. We do expect, however, in line with every FOMC voter, that the unemployment rate is likely to approach the vicinity of last cycle’s lows before the current phase ends. If that expectation is realized, there is a stretch of downward movement ahead (a good chunk of the 1.7% standard deviation, though July claimed 50 basis points of it) that has empirically been quite favorable for the S&P 500. The speed with which it covers the ground from here to 4% or below is unknown. Given the tremendous pent-up demand for labor, as evidenced by a record high job openings rate (Chart 3), the unemployment rate may come down much faster than it normally does. The level-and-direction analysis makes it clear that 5.9% and falling has provided an auspicious backdrop for equity investors, and the Fed’s more relaxed reaction function may allow the economy to run a little hotter than it normally would once unemployment falls below its natural rate. All in all, the empirical record of the relationship between the unemployment rate and equities suggests that stocks have room to run while the labor market improves. Earnings The unemployment rate may not be too low for equities to continue to rally, but is earnings growth too good for stocks’ own good? It doesn’t appear to be, given the historical interaction between forward one-quarter S&P 500 performance and the speed and acceleration of growth in trailing four-quarter earnings. We use trailing earnings because they exhibit extended trends that highly variable sequential changes in single-quarter data do not. Since 1948, trailing four-quarter operating earnings have experienced eleven complete double-digit declines from cycle peaks and eleven complete earnings growth phases, while beginning a new growth phase in the first quarter (Chart 4). Chart 4Steady Growth With Occasional Hiccups The chart shows that four-quarter earnings have grown in a pattern that features extended growth phases punctuated by concentrated declines that are occasionally severe. This pattern is the mirror image of the unemployment rate’s and S&P 500 earnings have been in a growth phase three out of every four quarters on the way to an annualized growth rate of 6.4%. Since P/E multiples are a mean-reverting series, stocks need to grow earnings to rise over time, but there is little difference in lagged S&P 500 returns when earnings are in growth or contraction mode (Chart 5). The disparity widens within each broad phase when we considered the growth rates – deceleration has been better for stock prices than acceleration within expansion phases, while a slowing rate of decline has been a tremendous catalyst when earnings are in a contraction phase. Chart 5More Money Is Made From Terrible To Bad Than From Good To Great To explore S&P 500 index performance during acceleration and deceleration phases within growth ranges, we repeated the unemployment rate analysis. The return disparities for different earnings ranges were not nearly as clear cut as they were for different unemployment ranges, but acceleration was good for near-term equity returns in the middle of the earnings growth distribution, while deceleration trumped acceleration at growth rates plus or minus three quarters of a standard deviation from the mean (Table 2). Table 2Headed Out Of The Earnings Sweet Spot The muddled empirical record does not point to a clear path for S&P 500 returns over the next few quarters. We assign a very low probability to a recession over the next year, virtually ensuring that the growth phase that began last quarter will continue. If actual earnings turn out to be somewhat close to the current consensus expectation, however, all subsequent quarters in this growth phase will be decelerating, and deceleration within growth phases (Table 2, circled three outcomes) has previously yielded below-average price returns. Trailing four-quarter earnings growth appears sustainable over the next year, however, and history is hardly sounding an alarm. Interest Rates We have already examined the relationship between moves in real 10-year Treasury yields and equity performance in a dedicated Special Report.1 The executive summary is that the level of real rates has exerted a greater influence on S&P 500 returns than their direction. The empirical evidence suggests that stocks generally outperform when real rates are rising, though they hit a wall once the real 10-yield exceeds estimated potential real GDP growth. They also underperform at the other extreme, as extremely negative real rates tend to be associated with dire economic conditions, but potentially frightening weakness is not a feature of today’s negative real-rate backdrop. Per the potential-GDP-rule-of-thumb, the nominal 10-year Treasury yield that would begin to crimp economic activity is around 4.5-5%, assuming potential GDP growth of 1.75-2% and annual inflation with a central tendency near 3%. It is very difficult to see the 10-year yield exceeding one-half of that threshold level in the next twelve months. Though a yield backup to 2% or above over the next year would likely have significant implications for relative returns within the S&P 500, we do not think it would spell the end of the equity rally. The bottom line, then, is that we do not believe that interest rates are at a level that makes equities especially vulnerable. Price-earnings multiples may well contract if real rates rise in line with our expectations, but we expect that earnings and earnings estimates would rise enough to offset the de-rating pressure. Investment Implications Mean reversion is a bedrock investment concept, and it helps explain why the level of variables that impact equity returns can be deceiving. When key variables reach extremes, the potential of an abrupt reversal increases. Financial markets are additionally forward discounting mechanisms and the rate of change – a variable’s “second derivative” – may offer more insight into its future path than its existing position. It is easy to see why investors typically favor direction over level when looking ahead. Level does not always take a back seat to direction, however, and we think a consideration of how level and direction interact is important when assessing the current landscape. Economic growth will surely slow from double or triple its long-run trend level, earnings will surely stop beating estimates by three or four times the maximum magnitude of the previous 32 quarters, nonfarm payrolls won’t expand by 900,000 every single month (though they may for much of the rest of this year) and a range of other variables won’t keep setting records. But deceleration from record highs will not necessarily spell the end of the rallies in risk assets. While important variables remain at elevated levels, equities and credit are likely to continue to generate excess returns. Extraordinary monetary and fiscal accommodation, combined with remarkably swift and successful action to blunt the threat of COVID-19, have carried financial markets for the last year-plus and we don’t think they’re finished yet. Doug Peta, CFA Chief US Investment Strategist dougp@bcaresearch.com Footnotes 1 Please see the September 24, 2018 US Investment Strategy Special Report, "When Will Higher Rates Hurt Stocks?", available at usis.bcaresearch.com.

Highlights Investors have grown enamored with online retailers (AMZN), payment processing companies (V, MA, PYPL, SQ), and social media companies (FB, SNAP). All three sectors are likely to experience headwinds over the next 12 months as life returns to normal following the pandemic. Looking further out, market saturation, increased competition, and heightened regulation all pose risks to these sectors. Internet companies in general, and social media firms in particular, will face increased scrutiny not just for their monopolistic practices, but for the mental harm they are causing young people. Just like cigarettes are heavily regulated due to their addictive qualities, the same could happen to social media. We think there is a 50/50 chance that governments will start restricting social media usage only to adults over the age of 18 by the end of the decade, a move that could decimate the sector. Global Growth Will Remain Above Trend Investors are worried about growth again. Globally, the number of Covid cases is on the rise due to the proliferation of the Delta variant (Chart 1). The ISM manufacturing index dropped to 59.5 in July, down from a high of 64.7 in March. Both of China’s manufacturing PMIs have fallen, with the new orders component of the Caixin index dipping below the 50 line. The European PMIs have also come off their highs (Chart 2). Chart 1Number Of Covid Cases On The Rise Globally Due To The Delta Variant Chart 2Manufacturing PMIs Are Off Their Highs     Growth concerns have registered in financial markets (Chart 3). After climbing to 1.74% in March, the US 10-year Treasury yield has fallen back to 1.22%. Cyclical equity sectors have underperformed defensives. Growth-sensitive currencies such as the Swedish krona and the Australian dollar have weakened. We are more upbeat about global growth prospects than the consensus. As the experience of the UK demonstrates, there is little will to impose lockdowns in countries with ample access to vaccines. Strict social distancing restrictions remain a fact of life in countries lacking adequate vaccine supplies. However, the situation should improve later this year as vaccine production increases (Chart 4). Chart 3Financial Markets Trim Growth Expectations Chart 4Over 10 Billion Vaccine Doses Will Be Produced This Year   Households in developed economies are sitting on US$5 trillion in excess savings, half of which reside in the United States (Chart 5). Inventories are at record low levels, which should support production over the coming quarters (Chart 6). Chart 5Households Flush With Excess Savings Chart 6Record Low Inventories Will Provide A Boost To Production     Chinese policy should turn more stimulative, as the recent cut to bank reserve requirements foreshadows. With credit growth back down to 2018 lows, policymakers can afford to give the economy some juice. The 6-month credit impulse has already turned up (Chart 7). From Goods To Services While global growth should remain well above trend for the next 12 months, the composition of that growth will shift in ways that could meaningfully affect equities. As Chart 8 illustrates, aggregate US consumption has returned to its pre-pandemic trend. However, spending on goods is 11% above trend while spending on services is still 6% below trend. Chart 7Chinese Policy Is Turning More Stimulative Chart 8The Divergence Between Goods And Services Spending   Households typically cut spending on durable goods during recessions, while services serve as the ballast for the economy. The opposite happened during the pandemic. As the global economy recovers, goods spending will slow while services spending will stay robust. This is critical for online retailers such as Amazon, which derive the bulk of their e-commerce revenue from selling goods. Even after its disappointing Q2 earnings report, analysts still expect Amazon to grow e-commerce sales by 17% in 2022 (Chart 9). Such a goal may be difficult to achieve, given that core US retail sales currently stand 13% above their trendline (Chart 10). Chart 9AAnalysts’ Great Expectations May Be Dashed (I) Chart 9BAnalysts’ Great Expectations May Be Dashed (II) Chart 10AUS Retail Spending Is Well Above Trend (I) Chart 10BUS Retail Spending Is Well Above Trend (II) Chart 11Screen Time Is Moderating If e-commerce spending slows, shares of payment processing companies could disappoint. Likewise, social media companies could suffer as people start going out more often. After spiking during the height of the pandemic, growth in data usage has returned to normal (Chart 11). Long-Term Risks Looking beyond the post-pandemic recovery, all three equity sectors face structural challenges that are not being fully discounted by investors. The first is market saturation. Close to three-quarters of US households have Amazon Prime accounts. Slightly over half have a Netflix account. Nearly 70% have a Facebook account. Google commands 92% of the internet search market. Together, Google and Facebook generate about 60% of all online advertising revenue. Competition is another challenge. Companies such as Amazon, Facebook, and Google dominate their respective markets. As they look for further growth, they will invariably invade each other’s turf. The result might benefit consumers, but it is unlikely to help the bottom line if it means more competitive pressures. Moreover, it is not just competition from within the tech industry that may disrupt incumbent firms. Consider payment processors. Like most other central banks, the Fed is planning to launch its own digital currency. Widely available, free-to-use Central Bank Digital Currencies (CBDCs) could thwart the ability of Visa and MasterCard to skim 2%-to-3% off of every transaction. Regulatory Pressures In recent years, tech companies have faced increased scrutiny over their alleged monopolistic practices. In contrast to Chinese tech firms, which have fallen under the thumb of the authorities, US companies have been able to evade harsh measures. Just last month, a US federal court judge dismissed a case filed by more than 40 state attorneys general arguing that Facebook’s acquisitions of Instagram and WhatsApp had harmed competition. In the past, evidence that companies were setting prices well above marginal costs could be used to build a case for anti-trust enforcement. Such cases are more difficult to argue today because so many online services are given away for free. Nevertheless, governments are likely to become more adept in pursuing regulatory actions. Rather than focusing simply on pricing policies, regulators are increasingly looking at the ways big tech companies use vendor data in the case of Amazon and user data in the case of Facebook and Google to maintain market dominance. Public contempt for tech companies is fueling a political backlash. According to a Gallup poll conducted earlier this year, only 34% of Americans held a favorable view of tech companies such as Amazon, Facebook, and Google, down from 46% in 2019; 45% had an unfavorable opinion, up from 33% in 2019. The shift in public sentiment over the past two years has been entirely driven by Independent and Republican voters, many of whom feel that tech companies are unfairly censoring their opinions (Table 1). The same poll revealed that the majority of Americans – including the majority of Republicans – now favor increased regulation of tech companies. Table 1American Views On Big Tech A Drug Worse Than Nicotine? Social media companies are among the most loathed within the tech sector. A Pew Research Center study conducted last year revealed that more than six times as many Americans had a negative opinion of social media as a positive one (Chart 12). The public’s disdain for social media is increasingly going beyond traditional concerns over privacy. As psychologists Jonathan Haidt and Jean Twenge recently argued in the New York Times, there is growing evidence that the pervasive use of social media is harming the mental health of the nation’s youth. The share of students reporting high levels of loneliness has more than doubled in both the US and abroad over the past decade (Chart 13). Chart 12Social Media Increasingly Vilified Chart 13Alone In The Crowd In 2019, the last year for which comprehensive data is available, nearly a quarter of girls between the ages of 12 and 17 reported experiencing a major depressive episode over the prior year, up from 12% in 2011 (Chart 14). Academic studies have shown that adolescents who use Facebook and Instagram frequently feel greater anxiety and unease than those who do not. Just like cigarettes are heavily regulated due to their addictive qualities, the same could happen to social media. Facebook and most other social media companies already restrict access to those under the age of 13, although enforcement is generally spotty. We assign a 50/50 chance that governments start restricting social media usage only to adults over the age of 18 by the end of the decade, a move that could decimate the sector. Priced For Perfection The seven companies in the three high-flying sectors mentioned in this report trade at 91-times forward earnings compared to the S&P 500’s aggregate multiple of 22. They also trade at an average price-to-sales ratio of 16 compared to 3.2 for the broader market (Chart 15). Chart 14The Rise In Depression Rates Coincided With Increased Social Media Usage Chart 15Trading At A High Multiple To Sales   Such valuations can be justified only if these companies grow earnings-per-share by nearly 30% per year over the next five years, as analysts currently expect (Chart 16). However, as noted above, that may be too high a hurdle to clear. Higher bond yields represent another threat to valuations. Growth stocks are much more sensitive to changes in discount rates than value stocks. Chart 17show that tech stocks have generally outperformed the S&P 500 over the past four years whenever bond yields were falling. We expect bond yields to rebound over the coming months, with the 10-year yield rising to 1.8% by early next year. Tech is likely to lag the market in that environment. Chart 16Long-Term Growth Estimates May Be Too Optimistic For These High-Fliers Chart 17Higher Bond Yields Could Hurt Tech Stocks   Trade Update Our long EM equity trade got stopped out last Tuesday before recouping some of its losses in subsequent days. We continue to expect EM stocks to bounce back later this year. That said, in keeping with this report, we see more upside for “traditional” EM sectors such as banks, industrials, energy, and materials than for EM tech (especially Chinese tech). Peter Berezin Chief Global Strategist pberezin@bcaresearch.com Global Investment Strategy View Matrix Special Trade Recommendations Current MacroQuant Model Scores

Highlights The DXY index appears to be following the seasonal pattern of strengthening in the summer and weakening towards year-end. In this context, the most attractive vehicles to play a decline in the dollar are the Scandinavian currencies over the longer term, and the yen in the very near term. Our composite attractiveness model ranks the US dollar and the NZ kiwi as the least attractive currencies, particularly on the basis of valuation. Our limit buy on long AUD/NZD was triggered at 1.05. Pessimism on the Aussie is becoming overdone, while the economy could stage a coiled spring rebound once vaccination rates improve. Feature Chart I-1Was Dollar Strength Seasonal? Since July 20, the DXY index has been consolidating its gains, and appears to be following the general seasonal pattern of strengthening in the summer, and eventually weakening towards year-end (Chart I-1). With this as a backdrop, it is instructive to revisit our attractiveness ranking, and highlight which currencies might benefit most from a dollar decline. Our framework is based on three major vectors –   the macroeconomic environment, valuation, and sentiment. Our macro vector tracks relative economic strength as measured by relative PMIs and real interest rate differentials. Other factors such as a country’s basic balance and external vulnerability are also considered. In our valuation vector, we consider a swathe of models including PPP, more high-frequency indicators such as our intermediate-term timing model, as well as longer-term models based on relative productivity trends. Finally, we also consider positioning to gauge if our view is mainstream or out of consensus. Using this framework, the most attractive vehicles to play a decline in the dollar are the Scandinavian currencies over the longer term, and the yen more near term, if rates remain well behaved. Meanwhile, the US dollar and the kiwi rank as the least attractive currencies, particularly on the basis of valuation (Chart I-2). Chart I-2An Attractiveness Ranking Of Currencies Macroeconomic Environment: Real Interest Rates Chart I-3The US Sports A Very Negative Real Yield On the short tenors, the US is among those sporting  the most negative real rates (Chart I-3). But what is interesting is that we know that there is a divergence in how various central banks are treating their inflation overshoot relative to the Federal Reserve. For example, both Norway and New Zealand have negative 2-year real rates, but their central banks are on track to lift short rates this year. However, the telegraphed messages from the Fed are that there will be no interest rate increases until 2023. This will push US real rates towards becoming more negative vis-à-vis other G10 countries. In our report titled Which Rates Matter For Currencies, we suggested that the recent decline in US Treasury yields should curtail strong inflows into US fixed income. This should ease upward pressure on the dollar. Macroeconomic Environment: Basic Balance Chart I-4Basic Balances Across The G10 The basic balance is one of the most important determinants of a currency’s attractiveness, simply because it captures the ebb and flow of demand for a country’s domestic assets. In a nutshell, the basic balance is the sum of the current account surplus and long-term investments. Trade surpluses underpin underlying demand for a country’s goods and services, while capital account surpluses suggest a country’s assets are under high demand. As such, persistent basic balance surpluses are usually associated with an appreciating currency and vice versa. There has been a sea change in the basic balances across the G10, a fact we highlighted in our recent report titled On The Fed Shift, And Balance Of Payments. One of those shifts involves Australia  seeing tremendous improvement in its basic balance surplus. In terms of rankings, Sweden sports the best basic balance surplus in the G10, followed by Australia and the euro area (Chart I-4). Meanwhile, the US ranks the worst in terms of basic balances, a big vulnerability for the currency. Macroeconomic Environment: External Debt A country’s external debt situation tends to only matter during crises. Therefore, in the current context of global fiscal and monetary stimulus, as well as generous Fed swap lines to assuage any dollar funding pressures abroad, external (especially USD) debt does not pose a significant threat for currencies. In an absolute sense, external debt as a share of GDP is highest in the UK, Switzerland, and Sweden (Chart I-5). However, what matters most often for vulnerability are net external assets rather than gross liabilities. Based on this measure, Japan, Norway, Canada, Switzerland and Sweden are the most attractive countries, based on net external assets (Chart I-6). Chart I-5External Debt In The G10 Chart I-6Net International Investment In The G10 Valuation: Purchasing Power Parity (PPP) Our PPP valuation model is our default in terms of evaluating a currency’s fair value, since by definition, it reveals price arbitration between any two countries.  Chart I-7The Dollar Is Expensive As we have documented, our model offers unique insight into a true PPP fair value, since it accounts for the fact that consumer price baskets tend to differ in composition from one country to the next. In order to get closer to an apples-to-apples comparison across countries, two adjustments are necessary. First, categorizing the consumer price index (CPI) into five major groups. In most cases, this breakdown captures 90% of the national CPI basket. This includes food, restaurants, and hotels (1), shelter (2), health care (3), culture and recreation (4), and energy and transportation (5). The second adjustment is to test the significance of individual price ratios, with the exchange rate as the dependent variable. This allows us to observe the most influential price ratios that help explain variations in the exchange rate. As a control strategy, we use a weighted average combination of the five groups to form a synthetic relative price ratio. If, for example, shelter is 33% in the US CPI basket, but 19% in the Swedish CPI basket, relative shelter prices will represent 26% of the combined price ratio. This allows for a uniform cross-country comparison, as opposed to using the national CPI weights. The results show the US dollar as overvalued, especially versus the Scandinavian currencies and the yen (Chart I-7). The results are based on the synthetic relative price ratio. Valuation: Intermediate-Term Timing Model (ITTM) Our ITTM is our favored model in the short term, because it gives signals with much higher frequency. Back in 2016, when we developed this indicator, it proved useful in helping global portfolio managers increase their Sharpe ratio in managing currency exposure. The idea was quite simple: For every developed world country, there were three key variables that influenced the near-term path of its exchange rate versus the US dollar: Interest Rate Differentials: We have elaborated at length that interest rate differentials are a key driver for currencies. Given that we get interest rates in real time, they are great inputs into any high-frequency model. Inflation Differentials: Inflation destroys the purchasing power of a currency, both in theory and practice (Chart I-8).  Assuming no transactional costs, the price of a dishwasher cannot be relatively high and rising in New York versus Manila. Either the US dollar needs to fall, the Philippine peso needs to rise, or a combination of the two has to occur to equalize prices across borders. Risk Factor: Exchange rates are risk assets. Ergo, the ebb and flow of risk aversion will have an impact on currencies, which is particularly the case for commodity exporters. We will be releasing a revamped version of our trading model in the coming weeks, incorporating results from ITTM. In a nutshell, our ITTM models have been a very good timing tool. And the signal today is to overweight JPY, AUD, SEK, and NOK in the G10 space (Chart I-9). Chart I-8Inflations and Currencies Chart I-9The Dollar Is Expensive Shorter Term Valuation: Long-Term Fair Value Model Chart I-10The Dollar Is Not Attractive Longer Term Our long-term FX models try to capture the movement in exchange rates over a business cycle (3-5 years, let’s say). Included in these models are much slower-moving variables like productivity differentials, and cumulative changes in the current account and basic balance. These models cover 22 currencies, incorporating both G10 and emerging market FX markets. We did an overhaul in these models this year, to account for rising Chinese productivity. Similar to our ITTM models, the longer-term valuation indicator favors the Scandinavian currencies, the yen, and the Aussie dollar (Chart I-10). Sentiment: Speculative Positioning The final consideration in our ranking is sentiment. In general, the dollar is a momentum currency and as such, you want to be long when bullish consensus and/or net speculative positioning is low and rising. Chart I-11 shows that the dollar has failed to break above its major trendlines, at the same time when bullish consensus on the dollar is rising (Chart I-12). This warns that a powerful countertrend reversal could be underway. Chart I-11The Dollar And Momentum Chart I-12The Dollar And Sentiment   According to CFTC data, the most shorted currencies are the Australian dollar and Japanese yen (Chart I-13). In our framework, these are the currencies slated to stage very powerful countertrend reversals, given we put the pandemic behind us. Chart I-13Everyone Is Long The Greenback Housekeeping Chart I-14AUD/NZD and Relative Rates Our long AUD/NZD position was triggered this week at 1.05. The messaging from the RBA and the RBNZ have been vastly different, whereby the former is cautious about the rising Delta variant infection rate, and the latter is focused on financial stability admist a bubbly housing market. On a relative policy basis, our bias is that the likelihood of rates adjusting higher than market expectations is higher in Australia than in New Zealand (Chart I-14). As we are  eventually going to put the virus behind us, underappreciated currencies such as the AUD could stage a mean-reversion rally.   Chester Ntonifor Foreign Exchange Strategist chestern@bcaresearch.com Trades & Forecasts Forecast Summary Core Portfolio Tactical Trades Limit Orders Closed Trades